This is an experimental project designed to evaluate the effectiveness of a nontraditional aftercare model (based on enhancement of individual social support systems) versus a standard aftercare model in controlling substance abuse and preventing HIV infection in two populations of substance abusers. It also examines the nature and significance of social support systems in drug using populations, i.e., what naturally occurring systems exist, how they are utilized, their relation to relapse and recovery. The data are derived from two sample sources: persons completing short term detoxification programs and drug involved offenders released from prison onto parole supervision. Both groups are randomly assigned to either the non-traditional aftercare service or the standard aftercare service and interviewed at three junctures: baseline (after detoxification/prior to released from incarceration), at six months post baseline and at one year post baseline. A sample of dropouts from each condition are also followed for comparison. The non-traditional aftercare service is a social support model which emphasizes enhancement of social functioning and utilization of social networks through peer counseling and support, provision of life support services and extensive referrals delivered in a community storefront setting. The standard aftercare services are basic group counseling services provided by professional staff in a traditional public health setting. Multiple regression techniques appropriate to panel data are employed.